Chinese Scientists at the China National Human Genome Centre (CHGC) in Shanghai are using SGI compute technology running a Mitrion-accelerated bioinformatics application to improve early disease diagnosis and discover new drugs to effectively treat disease caused by this parasite, such as Schistosoma japonicum or blood fluke.

Blood fluke has evolved for thousands of years and is one of the major infectious parasites to a wide range of hosts including primates, rodents, carnivores, and humans. Currently, existing drugs target disease caused by blood fluke found in South America. In the past few years the Schistosoma japonicum has been re-found in some lakes and rivers in the east and south China area. To study the parasite evolution, improve disease diagnosis in the very early stages, and develop more effective drugs to treat disease, Chinese bioscientists decided to discover the secrets of the blood fluke genome.

According to the announcement made by SGI, CHGC implemented the first phase of blood fluke genomics sequencing using the new SGI RASC Appliance and SGI InfiniteStorage 350 storage solution, installed in February. CHGC bioscientists achieved faster query times — up to 10 times faster — with the combination of SGI RASC (Reconfigurable Application-Specific Computing) technology, an accelerated version of BLAST-n software developed by Mitrionics, and the acclaimed SGI Altix server platform. Moreover, the InfiniteStorage 350 ensures that the data generated by the CHGC research is both protected and available to maximize analysis, outlines the report.

"Shortened time to results is critical to our success. With the blood fluke genome research, we have 300 million base pairs to study, and have 6-7 times more calculations in each step. The large shared memory and ease-of-use with the SGI RASC Appliance for BioInformatics enables our scientists to focus on achieving results faster and not spending valuable time on computer science," said Dr.Zhou, Deputy Director of Bioinformatics Department, CHGC.

The SGI RASC Appliance for Bioinformatics is a pre-configured solution that simplifies nucleotide sequence queries using BLAST-n. As outlined by SGI, The industrial-scale appliance addresses productivity problems in a range of bioinformatics environments — from those that serve thousands of users running BLAST queries against a single database, to others with smaller numbers of users running complex queries against databases that are hundreds of Gigabytes in size.

Using Mitrionics software and an accelerated version of NCBI BLAST-n, the SGI RASC Appliance for BioInformatics offloads genome sequencing workloads that typically run on Linux clusters. The Mitrion accelerated BLAST-n running on the SGI RASC Appliance for BioInformatics runs large queries up to 15 times faster than a single-core r nodes powered by AMD Opteron 8820 SE processors, and production runs of thousands of smaller queries by up to 60 times faster.

The SGI RASC Appliance for BioInformatics further increases throughput by executing multiple BLAST-n queries in parallel on multiple FPGAs.Furthermore, with up to 16 FPGAs in a single SGI RASC Appliance for BioInformatics, customers can achieve throughput equal to between 240 and 960 AMD Opteron cores without the solution complexity and system management overhead.

The CHGC also recognized that the research into the genome of the blood fluke would require them to run much more than BLAST-n. The scalable, general purpose nature of the SGI RASC Appliance for BioInformatics supported that capability by enabling the system to be upgraded to a total of 48-cores of Intel Itanium 2 processors, 128GB of shared memory and an InfiniteStorage 350 with 8TB of disk storage. The system's ease-of-use and portability of open source software running in the shared memory environment have proved crucial for the second phase of their research. The SGI RASC Appliance for BioInformatics is running Novell's SUSE® Linux Enterprise Server 10.